Trends in daily temperature and precipitation extremes over Georgia,1971–2010

I. Keggenhoff a,n, M. Elizbarashvili b, A. Amiri-Farahani c, L. King a

a Justus-Liebig-University Giessen, Department of Geography, Senckenbergstrasse 1, 35390 Giessen, Germanyb Ivane Javakhishvili Tbilisi State University, Department of Geography, 1, Chavchavdze Ave., 0179 Tbilisi, Georgiac University of California, Riverside, Department of Earth Sciences, 900 University Ave., Riverside, California 92521, United States of America

a r t i c l e i n f o

Article history:Received 3 December 2013Received in revised form23 April 2014Accepted 1 May 2014Available online 10 May 2014

Keywords:Daily temperature and precipitationClimate trendsClimate extreme indicesGeorgiaSouthern Caucasus

a b s t r a c t

Annual changes to climate extreme indices in Georgia (Southern Caucasus) from 1971 to 2010 arestudied using homogenized daily minimum and maximum temperature and precipitation series.Fourteen extreme temperature and 11 extreme precipitation indices are selected from the list of coreclimate extreme indices recommended by the World Meteorological Organization – Commission forClimatology (WMO-CCL) and the research project on Climate Variability and Predictability (CLIVAR) ofthe World Climate Research Programme (WCRP). Trends in the extreme indices are studied for 10minimum and 11 maximum temperature and 24 precipitation series for the period 1971–2010. Between1971 and 2010 most of the temperature extremes show significant warming trends. In 2010 there are13.3 fewer frost days than in 1971. Within the same time frame there are 13.6 more summer days and7.0 more tropical nights. A large number of stations show significant warming trends for monthlyminimum and maximum temperature as well as for cold and warm days and nights throughout thestudy area, whereas warm extremes and night-time based temperature indices show greater trends thancold extremes and daytime indices. Additionally, the warm spell duration indicator indicates a significantincrease in the frequency of warm spells between 1971 and 2010. Cold spells show an insignificantincrease with low spatial coherence. Maximum 1-day and 5-day precipitation, the number of very heavyprecipitation days, very wet and extremely wet days as well as the simple daily intensity index all showan increase in Georgia, although all trends manifest a low spatial coherence. The contribution of veryheavy and extremely heavy precipitation to total precipitation increased between 1971 and 2010,whereas the number of wet days decreases.& 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-SA

license (http://creativecommons.org/licenses/by-nc-sa/3.0/).

1. Introduction

The globally averaged surface temperature data shows a linearwarming trend of 0.85 1C [0.65 to 1.06 1C] during the period 1880–2012. The total increase between the average of the 1850–1900period and the 2003–2012 period is 0.78 1C [0.72 to 0.85 1C], basedon the single longest dataset available (IPCC, 2013). Climateextremes receive much attention as trends in extreme eventsreact more sensitively to climate change than mean climate, andtherefore have a more intense impact on natural and humansystems. (Katz and Brown, 1992; Easterling et al., 1997, 2000;Kunkel et al., 1999; New et al., 2006; IPCC, 2007; Aguilar et al.,2009). The IPCC (2012) stated that there is a high confidence

economic losses from weather- and climate-related disasters haveincreased during the last 60 years and will have greater impacts onsectors with closer links to climate, such as water, agriculture andfood security, whereas the highest fatality rates and economiclosses caused by hydro-meteorological induced disasters areregistered in developing countries. On a global scale, temperatureindices demonstrate a significant warming during the 20th cen-tury, citing the highest trends for the most recent periods and forminimum temperature indices. However, trends in extreme pre-cipitation illustrate a much lower spatial coherence, yet on a globalscale a significant wetting trend could be detected, whereas thenumber of consecutive dry days shows very different regionalchanges (Frich et al., 2002; Alexander et al., 2006). Since the 1990s,various regional studies pertaining to temperature and precipita-tion extreme indices have been conducted which provide strongevidence that global warming is related to significant changes intemperature and precipitation extremes. (Zhang et al., 2000;Manton et al., 2001; Peterson et al., 2002; Aguilar et al., 2005;Griffiths et al., 2005; Zhang et al., 2005b; Haylock et al., 2006;

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/wace

Weather and Climate Extremes

http://dx.doi.org/10.1016/j.wace.2014.05.0012212-0947/& 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-SA license (http://creativecommons.org/licenses/by-nc-sa/3.0/).

n Corresponding author. Tel.: þ49 641 99 36265; fax: þ49 641 99 36259.E-mail addresses: ina.keggenhoff@geogr.uni-giessen.de (I. Keggenhoff),

mariam.elizbarashvili@tsu.ge (M. Elizbarashvili),anahita.amirifarahani@email.ucr.edu (A. Amiri-Farahani),lorenz.king@geogr.uni-giessen.de (L. King).

Weather and Climate Extremes 4 (2014) 75–85

www.sciencedirect.com/science/journal/22120947www.elsevier.com/locate/wacehttp://dx.doi.org/10.1016/j.wace.2014.05.001http://dx.doi.org/10.1016/j.wace.2014.05.001http://dx.doi.org/10.1016/j.wace.2014.05.001http://crossmark.crossref.org/dialog/?doi=10.1016/j.wace.2014.05.001&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1016/j.wace.2014.05.001&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1016/j.wace.2014.05.001&domain=pdfmailto:ina.keggenhoff@geogr.uni-giessen.demailto:mariam.elizbarashvili@tsu.gemailto:anahita.amirifarahani@email.ucr.edumailto:lorenz.king@geogr.uni-giessen.dehttp://dx.doi.org/10.1016/j.wace.2014.05.001

Klein Tank et al., 2006; Alexander et. al, 2006, Skansi et al., 2013).Climate extreme indices have also been computed and analyzedfor projected future climate conditions based on simulations byGlobal Climate Models (GCMs) (Sillmann, 2013a and 2013b; Russoand Sterl, 2011; Tebaldi et al., 2006; Alexander and Arblaster,2009; Kharin et al., 2007; Orlowsky and Seneviratne, 2012).Extreme precipitation and extreme warm temperature are bothpredicted to increase while extreme cold temperatures willbecome more moderate. The results also suggest that climatemodels cannot overcome uncertainties for projecting futurechanges in extreme events, especially concerning the regionalscale with complex physiographic conditions. For that reason,analyzing changes in extreme climate events based on observationdata is essential. Developing and transition countries such asGeorgia face a number of political, financial and institutionalbarriers for a proper climate data monitoring system, includinglimitations on funding, technology and human resources (Page etal., 2004). After the collapse of the Soviet Union in the early 1990sthe number of meteorological stations in Georgia rapidlydecreased and the lack of station maintenance caused largemeasuring gaps. The quality and quantity of accessible climateseries still limit our understanding of the observed changes inclimate extremes in Georgia. So far, studies on changes in climateand climate extremes have been carried out based on monthlytemperature data and associated weather and climate phenomena,such as drought, hurricanes and frost (Elizbarashvili, 2007,2009, 2011, and 2012). Elizbarashvili (2013) discovered thatthe frequency of extremely hot months during the 20th centuryincreased, especially over Eastern Georgia, whereas extremely coldmonths decreased faster in the Eastern than in the Western region.In addition, highest rates for positive trends of mean annual airtemperature can be observed in the Caucasus Mountains. The aimof this study is to provide a better understanding of recent changesin the variability, intensity, frequency and duration of climateextreme events across Georgia by investigating trends in selecteddaily temperature and precipitation extreme indices between1971and 2010 using the software RClimDex 1.1 (Zhang and Yang, 2004).This is achieved by calculating a set of 25 ETCCDI climate extremeindices from homogenous daily weather data and estimatinglinear trends. Standardized anomalies of those trends are alsoinvestigated. The indices of temperature and precipitationextremes considered in the present study are recommended bythe Expert Team on Climate Change Detection Indices (ETCCDI).Section 2 of the current study describes the study area, the qualitycriteria of the data selection, quality control, homogenization, theclimate extreme indices and the analytical methods used in thisstudy. Section 3 presents and discusses the observed regional

trends for the analysis period 1971–2010 as averaged regionaltrends from anomalies and regional trends per decade for eachindex. Section 4 summarizes the conclusions.

2. Data and methods

2.1. Study area

Georgia is located in the Southern Caucasus between 41 and 441Nand 40 and 471E and covers an area of 69.700 km² (Figs. 1 and 2).It borders Russia to the North, Azerbaijan to the Southeast andArmenia and Turkey to the South. The topographic patternsthroughout Georgia are very diverse. The relief declines from theGreater Caucasus Range in the North, with an elevation range of1500–5000 m and the Lesser Caucasus with altitudes up to3500 m in the South towards Transcaucasia, which stretches fromthe Black Sea coast to the Eastern steppe. The Surami mountainchain with a maximum altitude of 1000 m connects the LesserCaucasus with the Greater Caucasus and divides Transcaucasia intoeastern and western lowlands (0–500 m). The Greater Caucasusrepresents an important climatic parting line towards Russia. Itprotects Transcaucasia from arctic high-pressure systems in winteroriginating from the Central Asian Region. The Southern Caucasusinhibits the summer heat from the Southeast. The Surami moun-tain chain avoids wet air masses circulating from the Black Seatowards the Caspian Sea causing high temperatures and humidclimate at the Western coast, continental climate in inner Trans-caucasia up to very dry climate with high temperatures in theeastern lowlands. Due to the complex annual large-scale circula-tion and diverse physiographic patterns, large spatial and temporaldifferences of temperature and precipitation over Georgia can beobserved. In general, the west of Georgia is characterized by mildwinters and hot summers with mean annual air temperatures of13 to 15 1C and high annual precipitation values (1200–2400 mm).The climate in eastern Georgia is continental with much lowerannual precipitation (500–600 mm in the lowlands) and a meantemperature between 10 and 13 1C. In the mountainous areasmean temperature covers a range of �5 to 10 1C and precipitationvaries from 800 to 1400 mm (World Bank, 2006).

2.2. Data, quality control and homogeneity testing

Daily minimum and maximum temperature and daily precipi-tation for 88 stations were kindly provided by the NationalEnvironmental Agency of Georgia (NEA). The analysis period1971–2010 was chosen to investigate recent change in extreme

Fig. 1. Stations with daily time series of temperature minimum (orange dots) and maximum (red dots) and precipitation for the period 1971–2010.

I. Keggenhoff et al. / Weather and Climate Extremes 4 (2014) 75–8576

temperature and precipitation and to optimize the number ofsignificant trends and data coverage throughout the study area.Data quality and homogeneity testing as well as the detection andadjustment of inhomogeneous time-series have been carried outusing the computer program RClimDex 1.1 and its softwarepackage RHtestV3 (Wang et al., 2010a) is accessible at: http://etccdi.pacificclimate.org. As a first step, temperature and precipi-tation time series with more than 20% missing values within theanalysis period 1971–2010 were excluded. Outliers in the timeseries have been identified and temporal consistency was testedaccording to Aguilar et al. (2003). Unphysical values, such asnegative precipitation or maximum temperature lower thanminimum temperature were set to missing values. Twenty-sixprecipitation series and 28 temperature minimum and maximumseries fulfilled all quality selection criteria. Observational climatedata can be influenced by different non-climatic effects, such asthe relocation of weather stations, land-use changes, adjustmentsin instruments and observational hours (Peterson et al., 1998;Aguilar et al., 2003). These effects result in inhomogeneity causinga shift in the mean of a time series, which could result in firstorder autoregressive errors. Breakpoints are not customarily docu-mented in metadata. If metadata does exist, it is often inaccessiblefrom archives, which has also been the case in this study. For thepresent study only homogenous time series were used (Appendix).In order to detect breakpoints in a data series the PenalizedMaximal F test was used, which is embedded in a recursive testingalgorithm (Wang, 2008a, 2008b). 28 minimum and maximumtemperature series showed 0.68 breakpoints on average. 10 mini-mum and 11 maximum temperature series were tested homo-genous. Daily precipitation series were tested using the softwareRHtest_dlyPrcp package recommended by Wang et al. (2010b).Two precipitation series were tested inhomogeneous and wereexcluded from the further analysis.

2.3. Climate extreme indices and analytical methods

The Expert Team (ET) and its predecessor, the CCl/CLIVAR/JCOMM Expert Team on Climate Change Detection and Indices(ETCCDI) defined 27 core climate extreme indices. The ETCCDIindices agreed by the international community aim to monitorchanges in “moderate” extremes and to enhance studies onclimate extremes using indices that are statistically robust, covera wide range of climates, and have a high signal-to-noise ratio(Zhang et al., 2011). The indices are calculated from daily tem-perature and precipitation data (Karl et al., 1999 and Peterson etal., 2001). From the core extreme indices 14 extreme temperature

and 11 extreme precipitation indices were selected for the presentstudy (Table 1). All trends for indices chosen have been calculatedannually using the software RClimDex 1.1.

Percentile indices were calculated using the standard referenceperiod 1971–2000 to make results easily comparable with otherstudies using the same reference period. During the trend estima-tion of the percentile-based indices, RClimDex 1.1 uses the boot-strapping approach to avoid possible bias within the referenceperiod associated with the existing inhomogeneity (Zhang et al.,2005a). Annual trends in temperature and extreme indices arecalculated for 10 temperature minimum, 11 temperature maximumand 24 precipitation series. During the calculation process particulardata requirements must be met in order to calculate index valuesusing RClimDex. An annual value is considered as incomplete ifmore than 15 days are missing in a year. A month will not becalculated when Z3 days are missing, a year will only be calculatedwhen all months are present. A percentile-based index will only becalculated if there is at least 70% data present within the referenceperiod. Additional ETCCDI-recommended standard criteria areapplied as described in the RClimDex user manual at: http://etccdi.pacificclimate.org/RClimDex/RClimDexUserManual.doc. Afterquality testing, RHtestV3 was used to test data homogeneity. Theserequirements for data completeness resulted in a final number ofextreme indices shown in Table 1, which outlines index abbre-viations, names, definitions, units and the number of stations forwhich each index has been computed. The trend calculations inthe present study were performed for the period 1971–2010 inorder to optimize significant results and data coverage through-out the study area. The magnitude of trends was calculated usingthe non-parametric Sen's slope estimator based on Kendall's tau(τ) (Sen, 1968). The statistical significance has been estimatedusing the Mann-Kendall test, whereas in the present study atrend was considered to be statistically significant if it was lessthan or equal to a level of 5% (Mann, 1945; Kendall, 1975). Resultsat the 25% level are also presented, as per Nicholls (2001), Theannual slopes of trends were converted into slope per decade. Inaddition, least-squared linear trends were calculated with respect toa 1971–2000 reference period for averaged anomaly series for eachindex, to provide an indication of annual fluctuations and forfurther scientific backup of the decadal trend or tendency calcu-lated. Trends have only been calculated for an index if less than 20%of the annual values were missing. The anomaly series werecalculated as follows:

xr;t ¼ ∑nt

i ¼ 1ðxi;t�xiÞ=nt ð1Þ

Fig. 2. Stations with daily time series of precipitation for the period 1971–2010.

I. Keggenhoff et al. / Weather and Climate Extremes 4 (2014) 75–85 77

http://etccdi.pacificclimate.orghttp://etccdi.pacificclimate.orghttp://etccdi.pacificclimate.org/RClimDex/RClimDexUserManual.dochttp://etccdi.pacificclimate.org/RClimDex/RClimDexUserManual.doc

where is the xr;t is the regionally averaged index at year t; xi;t is theindex for station i at year t; xiis the index mean at station i over theperiod 1971–2010; nt is the number of stations with data in year t.

In compliance with New et al. (2006), xi;t and xiwere standar-dized by dividing them by the standard deviation of the respectivestation to avoid the average series being dominated by stationswith high values and to make stations more comparable given thediverse climatic conditions in Georgia. The 10-year moving aver-age was also used to show the annual variation of climaticextremes within the analysis period.

3. Results and discussion

3.1. Trends in temperature extremes

Average regional trends of frost days (FD, summer days (SU),and tropical nights (TR) show significant warming (Figs. 3a, g,and j), whereas ice days (ID) show an insignificant but alsoincreasing trend (Fig. 3d). The trend per decade in the Georgiaaverage significant at the 5% level for the annual number of frostdays is �3.3 days/decade; for ice days it is �1.0 days/decade; forsummer days this trend is 3.4 days/decade and for tropical nightsit is 1.7 days/decade (Table 2). Figs. 4a–d displays a high spatialcoherence of the regional warming trends, particularly for frostdays (FD) and summer days (SU). There are no cooling trendssignificant at the 5% or 25% level. The warming trend for FD issignificant at the 5% (25%) level for 40% (80%) of all 10 stationsanalyzed; the significant warming trend for SU is achieved for 64%(91%) of the stations. The warming trend for ID is significant for 9%(27%) of 11 stations, whereas for TR it is significant for 20% (40%)stations (Table 2). From the averaged trends per decade overGeorgia, there are 13.3 fewer frost days and 4.0 fewer ice days in

the year 2010 than in the year 1971. In the same time period thereare 13.6 more summer days and 7.0 more tropical nights.

All averaged anomaly series between 1971 and 2010 in theannual number of cold nights (TN10p), warm nights (TN90p), colddays (TX10p) and warm days (TX90p) show significant warmingtrends, whereas temperature minimum-based indices show largertrends (Table 2). No significant cooling trend is apparent (Fig. 4c-f). For TN10p the warming trend is significant at the 5% (25%)level for 70% (80%) of all 10 stations analyzed. The trend perdecade averaged over Georgia in the annual number of cold days is�1.0 days/decade. For TX10p a warming trend at a slightly lowerrate of �0.8 days/decade can be observed (Table 2). The trend perdecade in the Georgia-average annual number of warm nightsamounts to 1.9 days/decade, whereas the trend for warm daysaccounts for 1.2 days/decade. The considerable warming trend forTN90p is significant at the 5% (25%) level for 80% (80%) out of 10stations. TX90p shows a warming trend with lower magnitudethan TN90p, which is also indicated by the standardized anomaliesseries in Figs. 3h and i. 45% (82%) of the stations are decidedlyincreasing for TX90p.

The cold spell duration indicator (CSDI) shows an insignificantdecreasing trend at a rate of �0.1 days/decade. However, thewarm spell duration indicator (WSDI) shows a strong significantwarming trend in the regional averages, whereas the spatialcoverage of significant trends for WSDI is much higher than forCSDI (Figs. 4g and h). The moving average of the standardizedanomaly series of WSDI shows a pronounced fluctuation between1971 and 2010 (Fig. 3n). There is a steady trend until the mid-1990s and a stronger increasing trend until 2010. The mean trendover Georgia significant at the 5% (25%) level for WSDI isconsiderably positive at a rate of 1.7 days/decade for 55% (100%)of all stations. The Georgia-averaged trend for CSDI is significantonly at the 25% and amounts to �0.1 days/decade (Table 2),whereas the trend of the anomaly series is weak and almost even

Table 1ETCCDI temperature and precipitation indices selected for this study with index names, definitions, units and the number of stations per index for which trends werecomputed for the period 1971–2010. The list of all core ETCCDI indices is given at http://etccdi.pacificclimate.org/list_27_indices.shtml.

ID Index name Definitions Units Number ofstations

SU Summer days Annual count when TX(daily maximum)425 1C days 11FD Frost days Annual count when TN(daily minimum)o0 1C days 10ID Ice days Annual count when TX(daily maximum)o0 1C days 11TR Tropical nights Annual count when TN(daily minimum)420 1C days 10TXx Max Tmax Monthly maximum value of daily maximum temp 1C 11TNx Max Tmin Monthly maximum value of daily minimum temp 1C 10TXn Min Tmax Monthly minimum value of daily maximum temp 1C 11TNn Min Tmin Monthly minimum value of daily minimum temp 1C 10TN10p Cool nights Percentage of days when TNo10th percentile of 1971–2000 days 10TX10p Cool days Percentage of days when TXo10th percentile of 1971–2000 days 11TN90p Warm nights Percentage of days when TN490th percentile of 1971–2000 days 10TX90p Warm days Percentage of days when TX490th percentile of 1971–2000 days 11WSDI Warm spell duration indicator Annual count of days with at least 6 consecutive days when TX490th percentile of 1971–

2000days 11

CSDI Cold spell duration indicator Annual count of days with at least 6 consecutive days when TNo10th percentile of 1971–2000

days 10

Rx1day Max 1-day precipitation amount Monthly maximum 1-day precipitation mm 24Rx5day Max 5-day precipitation amount Monthly maximum consecutive 5-day precipitation mm 24SDII Simple daily intensity index Annual total precipitation divided by the number of wet days (defined as PRCP4¼1.0mm)

in the yearmm/day

24

R1mm Number of wet days Annual count of days when PRCP4¼1 mm days 24R10mm Number of heavy precipitation days Annual count of days when PRCP4¼10 mm days 24R20mm Number of very heavy precipitation

daysAnnual count of days when PRCP4¼20 mm days 24

R95p Very wet days Annual total PRCP when RR495th percentile of 1971–2000 mm 24R99p Extremely wet days Annual total PRCP when RR499th percentile of 1971–2000 mm 24CDD Consecutive dry days Maximum number of consecutive days with RRo1 mm days 24CWD Consecutive wet days Maximum number of consecutive days with RR4¼1 mm days 24PRCPTOT Annual total wet-day precipitation Annual total PRCP in wet days (RR4¼1 mm) mm 24

I. Keggenhoff et al. / Weather and Climate Extremes 4 (2014) 75–8578

http://etccdi.pacificclimate.org

(Fig. 3m). All warm extremes (TN90p and TX90p, WSDI) showhigher trend magnitudes than those for cold extremes (TN10pand TX10p, CSDI), implying asymmetric changes in lower- and

upper-tail extremes causing a rising temperature variance(Figs. 3b, c, h, i, m and n). This finding is in agreement with earlierstudies (Klein Tank and Kon̈nen, 2003; Zhang et al., 2005b;

Fig. 3. Averaged regional trends for temperature indices: FD, TN10p, TX10p, ID, TNN, TXN, SU25, TN90p, TX90p, TR, TNX, TXX CSDI and WSDI for the period 1971–2010. R isthe correlation coefficient of the linear trend. The dashed line is the 10-year moving average. (a) Frost days (FD). (b) Cool nights (TN10p). (c) Cool days (TX10p). (d) Ice days(ID). (e) Monthly minimum of Tmin (TNN). (f) Monthly minimum of Tmax (TXN). (g) Summer days (SU). (h) Warm nights (TN90p). (i) Warm days (TX90p). (j) Tropical nights(TR). (k) Monthly maximum of Tmin (TNX). (l) Monthly maximum of Tmax (TXX). (m) Cold spell duration indicator (CSDI). (n) Warm spell duration indicator (WSDI).

I. Keggenhoff et al. / Weather and Climate Extremes 4 (2014) 75–85 79

Table 2Georgia-averaged trends in temperature extreme indices for the period 1971–2010. Mean trends significant at the 5% level are set in bold. The percentage of stations withnegative and positive trends for all stations and stations significant at the 5 and 25% level for each index.

Temperature index Mean Range % of stations with neg. trends % of stations with pos. trends

sig. at 5% sig. at 25% all sig. at 5% sig. at 25% all

FD [days] �3.3 �4.5 to -0.2 40 80 100 0 0 0ID [days] �1.0 �2.3 to 0.3 9 27 82 0 0 18TN10p [days] �1.0 �1.7 to 0.1 70 80 90 0 0 10TX10p [days] �0.8 �1.2 to 0.1 18 64 91 0 0 9TNn [1C] 0.4 0.0 to 1.0 0 0 0 20 40 100TXn [1C] 0.2 �0.2 to 0.6 0 0 18 0 27 82SU [days] 3.4 0.7 to 6.2 0 0 0 64 91 100TR [days] 1.7 0.0 to 7.1 0 0 0 20 40 100TN90p [days] 1.9 �0.1 to 3.3 0 0 10 80 80 90TX90p [days] 1.2 0.4 to 2.6 0 0 0 45 82 100TNx [1C] 0.3 �0.1 to 0.5 0 0 10 30 40 90TXx [1C] 0.3 �0.1 to 0.8 0 0 9 18 55 91WSDI [days] 1.7 1.6 to 6.7 0 0 0 55 100 100CSDI [days] �0.1 �1.0 to 0.3 10 20 60 0 0 40

Fig. 4. Trends per decade in the (a) annual number of frost days (FD), (b) annual number of summer days (SU), (c) cold nights (TN10p), (d) warm nights (TN90p), (e) cold days(TX10p), (f) warm days (TX90p), (g) cold spell duration indicator (CSDI) and (h) warm spell duration indicator (WSDI) for the period 1971–2010. Positive trends arerepresented by upward, negative trends by downward triangles. The red color indicates warming trends, blue indicates cooling trends. Light blue and red triangles indicatetrends not significant at the 5% level, white triangles are trends not significant at the 25% level. (a) Frost days (FD). (b) Summer days (SU). (c) Cool nights (TN10p). (d) Warmnights (TN90p). (e) Cool days (TX10p). (f) Warm days (TX90p). (g) Cold spell duration indicator (CSDI). (h) Warm spell duration indicator (WSDI)

I. Keggenhoff et al. / Weather and Climate Extremes 4 (2014) 75–8580

Moberg et al., 2006) demonstrating that warming since the 1970 sis rather caused by the increases of warm extremes than thedecrease of cold extremes. All maximum and minimum tempera-ture indices show an increasing trend (Fig. 3 e, f, k and l), while thewarmest night temperature (TNx) showed the only significanttrend rate of 0.3 days/decade averaged over Georgia. The coldestnight temperature (TNn) showed a warming rate of 0.4 1C/decade.For both indices the magnitude and the number of stations withsignificant trends were larger than for the warmest day tempera-ture (TXn) and the coldest day temperature (TXx). The warmingtrend for TXn accounts for 0.2 1C/decade (Table 2). For TXx thepositive trend is at a rate of 0.3 1C/decade (Figs. 3k and l). Trends intemperature minimum indices (FD, TR, TNx, TNn, TN10p andTN90p) are larger than those of their corresponding maximumextremes (SU, ID, TXn, TXx, TX10p and TX90p) (Figs. 3a� l). Theseobservations correspond to findings in previous studies relatingsignificant changes in temperature extremes with warming. In thecase of minimum temperature indices, most increasing trends andhigh significance throughout the study areas were observed,denoting that warming trends for night-time indices are largerthan for daytime indices (Manton et al., 2001; Peterson et al.,2002; Aguilar et al., 2005; Griffiths et al., 2005; Klein Tank andKon̈nen, 2003; Klein Tank et al., 2006; New et al., 2006).

3.2. Trends in precipitation extremes

The significance of trends in precipitation extremes from 1971–2010 is much lower compared to temperature extremes. A largeproportion of stations experienced an increase for the number ofvery heavy precipitation days, very wet and extremely wet days,maximum 5-day precipitation and the simple daily intensityindex, although the percentage of significant trend values wasvery low (Table 3). A negative Georgia-averaged trend wasdetected for the number of wet days (R1 mm) and the numberof heavy precipitation (R10 mm), yet the magnitude of trends forR10 mm was very weak (Fig. 5b).

None of the precipitation indices showed significance at the 5%level. Thus, all station trends in the present study were indicatedfor the 5 and 25% level. The Georgia-averaged trend of total wet-day precipitation (PRCPTOT) was insignificant at a rate of 7.9 days/decade, although the largest proportion of stations show asignificant negative trend at the 5% level for 12% for all 24 stationsanalyzed and the station trends show a low spatial coherencedistributed over Georgia and (Table 3 and Fig. 6a). In addition, thetrend of the standardized anomaly is very week. It shows adecreasing trend until the end of the 2000s and increases againat a higher magnitude (Fig. 5a). This is also the instance for thenumber of wet days, heavy precipitation days and consecutive wet

days (CWD). However, R1mm and R10mm indicated a decreasingtrend of �0.8 days/decade and �0.1 days/decade, respectively,whereas the trend for R10 mm is almost even (Figs. 5b and d).The negative trend for R1mm is significant at the 5% (25%) level for18% (29%) of all stations analyzed. R10mm is significant at the 5%(25%) level for only 12% (12%) of the stations and those withpositive trends account for 12% (24%). For the number of veryheavy precipitation days (R20 mm) the Georgia-averaged trend of0.3 days/decade was observed. It is significant at the 5% (25%) levelfor 6% (18%) of 24 stations (Table 3).

The increase of very wet days (R95p) and extremely wet days(R99p) is continuous and stable (Figs. 5e and f). The Georgia-averaged trend for R95p accounts for 7.5 mm/decade and for R99p6.3 mm/decade, respectively, whereas the positive trend for R95pis significant at the 5% (25%) level for 0% (29%) and the trend forR99p for 18% (35%) of the 24 stations analyzed (Table 3). Theamplified contribution of extreme precipitation to total precipita-tion was studied by employing the example of Groisman et al.(1999). The fraction of the annual averaged precipitation amountdue to very wet days increased by 23% for the period 1971–2010(range 14% to 33%) (Fig. 7a). The average contribution of extremelywet days increased at a rate of 7% (range 2–12%) (Fig. 7b). Thisamplified response of extreme precipitation to total precipitationis parallel to the increase of extreme precipitation events in thecontext of global changes, whereas the tendency for an increase inheavy daily precipitation events can also be found in regions withprojected decreases of total precipitation (Groisman et al., 2005;Alexander et al., 2006; IPCC, 2007, 2012).

The maximum 1-day precipitation (RX1day) indicated aninsignificant increasing trend of 0.6 mm/decade significant at the5% (25%) level for 11% (21%) of all stations, although it has a verylow spatial coherence and even shows a slightly negative tendencyfor the standardized anomaly series (Fig. 5g). However, maximum5-day precipitation (Rx5day) demonstrated a positive trend at arate of 2.0 mm/decade between 1971 and 2010, which is signifi-cant at the 5% (25%) level for 11% (16%) (Table 3).

The Georgia-averaged simple daily intensity index (SDII) indi-cates the most considerable and continuous trends (Fig. 5i) of allprecipitation indices. It has experienced a positive trend of0.1 mm/day/decade significant at the 5% (25%) level for 6% (24%)of the 24 stations analyzed (Table 3). The trend for consecutive drydays (CDD) is positive at a rate of 0.1 days/decade, which is inaccordance with the projections for the increase in dryness for theMediterranean area and central Europe (IPCC, 2012).The trend issignificant at the 5% (25%) level for 6% (12%) of the stations, but thenumber of stations with significant trends has a very low spatialcoherence and the proportion of stations with negative andpositive trends is almost even (Table 3 and Fig. 5k). CWD showed

Table 3Georgia-averaged trends in precipitation extreme indices for the period 1971–2010. Mean trends are significant at the 25% level. The percentage of stations with negative andpositive trends for all stations and stations significant at the 5 and 25% level for each index.

Precipitation index Mean Range % of stations with neg. trends % of stations with pos. trends

sig. at 5% sig. at 25% all sig. at 5% sig. at 25% all

PRCPTOT [mm] 7.9 �58.0 to 56.9 12 29 41 0 29 59R1mm [days] �0.8 �12.4 to 3.1 18 29 71 6 12 29R10mm [days] �0.1 �2.1 to 2.1 12 24 47 6 12 53R20mm [days] 0.3 �1.2 to 2.5 0 18 24 6 18 76R95p [mm] 7.5 �37.4 to 59.0 6 18 29 0 29 71R99p [mm] 6.3 �14.2 to 28.4 0 6 35 18 35 65Rx1day [mm] 0.6 �8.4 to 7.4 5 16 42 11 21 58Rx5day [mm] 2.0 �9.7 to 9.7 0 0 32 11 16 68CWD [days] 0.0 �2.1 to 0.9 12 12 59 6 12 41CDD [days] 0.1 �1.8 to 3.9 0 6 59 6 12 41SDII [mm/day] 0.1 �0.3 to 0.6 6 12 24 6 24 76

I. Keggenhoff et al. / Weather and Climate Extremes 4 (2014) 75–85 81

also a very weak trend of the standardized anomaly series and iseven at a rate of 0.0 days/decade, which is significant at the 5%(25%) level for 6% (12%) of the stations analyzed (Table 3).

4. Conclusions

This study analyzed changes in temperature and precipitationextreme indices in Georgia based on daily minimum and maximumtemperature and precipitation series for the period 1971–2010. Thedata was quality controlled and the temperature series was homo-genized using the software RClimDex 1.1. Due to moderate dataquality many station series had to be rejected during the data

quality assessment and computation process of the ETCCDI extremeindices, specifically, series of percentile-based precipitation indices.A large number of temperature series were excluded from the studydue to inhomogeneity. Nevertheless, the study could improve theunderstanding of recent changes in the variability, intensity, fre-quency and duration of climate extreme events over Georgia.Following changes for temperature and precipitation indices wereobserved throughout the study area:

� There are significant warming trends in the Georgia-average forSU, FD, TR, TXx, TN90p, TX90p, TN10p, TX10p and WSDIbetween 1971 and 2010.

Fig. 5. Averaged regional trends for precipitation indices: PRCPTOT, R10 mm, R20 mm, R1 mm, R95p, R99p, RX1day, RX5day, SDII, CWD and CDD for the period 1971–2010. Ris the correlation coefficient of the linear trend. The dashed line is the 10-year moving average. (a) Annual total wet-day precipitation (PRCPTOT). (b) Number of heavyprecipitation days (R10 mm). (c) Number of very heavy precipitation days (R20 mm). (d) Number of wet days (R1 mm). (e) Very wet days (R95p). (f) Extremely wet days(R99p). (g) Maximum 1-day precipitation (Rx1day). (h) Maximum 5-day precipitation (Rx5day). (i) Simple daily intensity index (SDII). (j) Consecutive wet days (CWD).(k) Consecutive dry days (CDD).

I. Keggenhoff et al. / Weather and Climate Extremes 4 (2014) 75–8582

� No significant cooling trend was found over Georgia for alltemperature indices.

� Most of the stations show significant warming trends for TNx,TNn TXx and TXn.

� The magnitude of trends for night-time indices (FD, TR, TNx,TNn, TN90p, and TN10p) is more pronounced than those fordaytime (SU, ID, TXx, TXn, TX90p, and TX10p).

� Averaged trends show “asymmetric” changes in temperatureextremes, indicating a more pronounced increase in warmextremes than a decrease in cold extremes.

� Although almost all extreme precipitation indices, such asR20mm, R95p, R99p, Rx1day, Rx5day and SDII indicate anincreasing number and more intense extreme precipitation

events over Georgia during 1971–2010, the changes were notstatistically significant.

� At the same time the number of wet and heavy precipitationdays (R1mm, R10mm) decreased

� The contribution of extreme precipitation to total precipitationincreased between 1971 and 2010.

Despite the small number of homogenous minimum andmaximum temperature series the study could present a highproportion of significant station and Georgia-averaged trends.The study provided evidence that during the last 40 years Georgiawas particularly affected by warm extremes based on night-timeindices rather than by cold extremes based on day-time indices.

Fig. 6. Annual trends per decade in the (a) annual total wet-day precipitation (PRCPTOT), (b) simple daily intensity index (SDII), (c) maximum 1-day precipitation amount(Rx1day), (d) number of very heavy precipitation days (R20 mm), (e) very wet days (R95p) and (f) extremely wet days (R99p) for the period 1971–2010. Positive trends arerepresented by upward, negative trends by downward triangles. The red color indicates drying trends, blue indicates wetting trends. Light blue and red triangles indicatetrends not significant at the 5% level, white triangles are trends not significant at the 25% level. (a) Annual total wet-day precipitation (PRCPTOT). (b) Simple daily intensityindex (SDII). (c) Maximum 1-day precipitation (Rx1day). (d) Number of very heavy precipitation days (R20 mm). (e) Very wet days (R95p). (f) Extremely wet days (R99p).

Fig. 7. Contribution of extreme precipitation to total precipitation. Averaged regional indices series show the fraction of precipitation due to (a) very wet days (R95ptot%)and (b) extremely wet days (R99ptot%). R is the correlation coefficient of the linear trend. The dashed line is the 10-year moving average.

I. Keggenhoff et al. / Weather and Climate Extremes 4 (2014) 75–85 83

Particularly warm days and nights and the warm spell durationindicator showed pronounced warming during 1971 and 2010indicating that an increase in the frequency, intensity and durationof temperature extremes can be expected in future. However, thesparse number of temperature stations made it impossible tomake reliable assumptions on trend patterns over Georgia. Thelow significance and spatial coherence of trends on precipitationextremes made it also very difficult to detect to provide significantevidence for spatio-temporal changes in extreme precipitationindices during 1971 and 2010. In particular, precipitation patternsover Georgia are very complex, influenced by the diverse topo-graphy and large-scale circulation. Thus, a study on the relationbetween large-scale circulation and changes in extreme tempera-ture and precipitation is necessary with higher temporal resolu-tion (seasonal or monthly) and at smaller spatial scales (e.g.regional and local). A further study on annual and seasonalchanges in temperature over Georgia comparing different analysisperiods is already in preparation. In addition, future analyses areplanned to study the relationship between changes in atmosphericcirculation and their contribution to the observed trends intemperature and precipitation extremes.

Acknowledgments

This study was made possible by the research grant “Interna-tional Postgraduate Studies in Water Technologies (IPSWaT)” ofthe International Bureau, Federal Ministry of Education andResearch. Additional support for fieldwork was provided by theGerman-Georgian project Amies (Analysing multiple interrelation-ships between environmental and societal processes in mountai-nous regions of Georgia) of the Volkswagen Stiftung. We greatlyappreciate the valuable suggestions by the reviewers for improv-ing our paper. Special thanks are due to the National Environ-mental Agency of Georgia (NEA) for its data contribution.

Appendix

See the appendix Table A1

References

Aguilar, E., Aziz Barry, A., Brunet, M., Ekang, L., Fernandes, A., Massoukina, M.,Mbah, J., Mhanda, A., do Nascimento, D.J., Peterson, T.C., Thamba Umba, O.,Tomou, M., Zhang, X., 2009. Changes in temperature and precipitation extremesin western central Africa, Guinea Conakry, and Zimbabwe 1955–2006.J. Geophys. Res. 114, D02115, http://dx.doi.org/10.1029/2008JD011010.

Aguilar, E., Peterson, T.C., Ramírez Obando, P., Frutos, R., Retana, J.A., Solera, M.,González Santos, I., Araujo, R.M., Rosa García, A., Valle, V.E., Brunet India, M.,Aguilar, L., Álvarez, L., Bautista, M., Castañón, C., Herrera, L., Ruano, R., Siani, J.J.,Hernández Oviedo, G.I., Obed, F., Salgado, J.E., Vázquez, J.L., Baca, M., Gutíerrez,M., Centella, C., Espinosa, J., Martínez, D., Olmedo, B., Ojeda Espinoza, C.E.,Haylock, M., Núñez, R., Benavides, H., Mayorga, R., 2005. Changes in precipita-tion and temperature extremes in Central America and northern SouthAmerica, 1961–2003. J. Geophys. Res. 110, D23107, http://dx.doi.org/10.1029/2005JD006119.

Aguilar, E., Auer, I., Brunet, M., Peterson, T.C., Wieringa, J., 2003. Guidelines onClimate Metadata and Homogenization. World Meteorological Organization,Geneva, Switzerland p. 55 (WMO-TD No. 1186, WCDMP No. 53).

Alexander, L.V., Arblaster, J.M., 2009. Assessing trends in observed and modeledclimate extremes over Australia in relation to future projections. Int. J. Climatol.29, 417–435, http://dx.doi.org/10.1002/joc.1730.

Alexander, L.V., Zhang, X., Peterson, T.C., Caesar, J., Gleason, B., Klein Tank, A.M.G.,Haylock, M., Collins, D., Trewin, B., Rahimzadeh, F., Tagipour, A., Rupa Kumar, K.,Revadekar, J., Griffiths, G., Vincent, L., Stephenson, D.B., Burn, J., Aguilar, E.,Brunet, M., Taylor, M., New, M., Zhai, P., Rusticucci, M., Vazquez-Aguirre, J.L.,2006. Global observed changes in daily climate extremes of temperature andprecipitation. J. Geophys. Res. 111, D05109, http://dx.doi.org/10.1029/2005JD006290.

Easterling, D.R., Horton, B., Jones, P.D., Peterson, T.C., Karl, T.R., Parker, D.E., Salinger,M.J., Razuvayev, V., Plummer, N., Jamason, P., Folland, C.K., 1997. Maximum andminimum temperature trends for the globe. Science 18, 364–367, http://dx.doi.org/10.1126/science.277.5324.364.

Easterling, D.R., Meehl, G.A., Parmesan, C., Changnon, S.A., Karl, T.R., Mearns, L.O.,2000. Climate extremes: observations, modeling, and impacts. Science 289,2068–2074, http://dx.doi.org/10.1126/science.289.5487.2068.

Elizbarashvili, E.Sh., Tatishvili, M.R., Elizbarashvili, M.E., Elizbarashvili, Sh.E.,Meskhiya, R.Sh., 2013. Air temperature trends in Georgia under global warmingconditions. Russ. Meteorol. Hydrol. 38, 234–238, http://dx.doi.org/10.3103/S1068373913040043.

Elizbarashvili, E.Sh., Varazanashvili, O.Sh., Tsereteli, N.S., Elizbarashvili, M.E., Eliz-barashvili, Sh.E., 2012. Dangerous fogs on the territory of Georgia. Russ.Meteorol. Hydrol. 37, 106–111, http://dx.doi.org/10.3103/S1068373912020057.

Elizbarashvili, E.Sh., Varazanashvili, O.Sh., Elizbarashvili, M.E., Tsereteli, N.S., 2011.Light frosts in the freeze-free period in Georgia. Russ. Meteorol. Hydrol. 36,399–402, http://dx.doi.org/10.3103/S1068373911060069.

Elizbarashvili, E.Sh., Meskhiya, R.Sh., Elizbarashvili, M.E., Megrelidze, L.D.,Gorgisheli, V.E., 2009. Frequency of occurrence and dynamics of droughts in20th century were studied based on the observational materials of 20meteorological stations of Eastern Georgia. Russ. Meteorol. Hydrol. 34,401–405, http://dx.doi.org/10.3103/S1068373909060107.

Elizbarashvili, E.Sh., Meskhiya, R.Sh., Elizbarashvili, M.E., 2007. Dynamics ofOccurrence Frequency of Extreme Anomalies of Monthly Mean Air Temperaturein Georgia in the 20th Century and Its Effect on Precipitation and on the RiverWater Discharge. Russ. Meteorol. Hydrol. 1, 71–74, http://dx.doi.org/10.3103/S1068373907010116.

Frich, P., Alexander, L.V., Della-Marta, P., Gleason, B., Haylock, M., Klein Tank, A.M.G.,Peterson, T., 2002. Observed coherent changes in climatic extremes during thesecond half of the twentieth century. Clim. Res. 19, 193–212, http://dx.doi.org/10.3354/cr019193.

Griffiths, G.M., Chambers, L.E., Haylock, M.R., Manton, M.J., Nicholls, N., Baek, H.-J.,Choi, Y., Della-Marta, P.M., Gosai, A., Iga, N., Lata, R., Laurent, V., Maitrepierre, L.,

Table A1

Stationname

WMOcode

Northlatitude

Eastlongitude

Altitude[m]

Analyzedparameter

Zemo Azhara 37196 43.10 41.73 952 Tmax/Precipitation

Zugdidi 37279 42.50 41.88 117 PrecipitationLentekhi 37295 42.77 42.72 731 Tmax/

PrecipitationAmbrolauri 37308 42.52 43.13 544 Tmax/

PrecipitationPoti 37379 42.12 41.70 1 PrecipitationSamtredia 37385 42.18 42.37 26 TminKutaisi 37395 42.20 42.60 116 Tmax/

PrecipitationSachkere 37403 42.35 43.40 455 PrecipitationMta-Sabueti 37409 42.00 43.50 1246 Tmax/

PrecipitationKhashuri 37417 42.00 43.60 690 Tmin/

PrecipitationPasanauri 37432 42.35 44.70 1064 Tmin/Tmax/

PrecipitationKobuleti 37481 41.87 41.77 7 TminBatumiAirport

37484 41.63 41.60 32 Precipitation

Khulo 37498 41.63 42.30 946 Tmax/Precipitation

Abastumani 37503 41.72 42.83 1265 Tmin/Precipitation

Ahalcihe 37506 41.63 42.98 982 PrecipitationGoderdziPass

37507 41.60 42.50 2025 Tmin

Borjomi 37515 41.83 43.38 794 Tmin/Precipitation

Gori 37531 41.98 44.12 590 PrecipitationTsalka 37537 41.60 44.07 1458 PrecipitationTbilisi 37549 41.75 44.77 427 PrecipitationTelavi 37553 41.93 45.38 562 Tmin/

PrecipitationKvareli 37563 41.97 45.83 449 Tmin/Tmax/

PrecipitationAkhalkalaki 37602 41.40 43.47 1716 Tmin/

PrecipitationParavani 37603 41.48 43.87 2100 Tmax/

PrecipitationBolnisi 37621 41.45 44.55 534 Tmax/

PrecipitationDedopliskaro 37651 41.50 46.10 800 Tmax/

Precipitation

I. Keggenhoff et al. / Weather and Climate Extremes 4 (2014) 75–8584

http://dx.doi.org/10.1029/2008JD011010http://dx.doi.org/10.1029/2008JD011010http://dx.doi.org/10.1029/2008JD011010http://dx.doi.org/10.1029/2005JD006119http://dx.doi.org/10.1029/2005JD006119http://dx.doi.org/10.1029/2005JD006119http://dx.doi.org/10.1029/2005JD006119http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref3http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref3http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref3http://dx.doi.org/10.1002/joc.1730http://dx.doi.org/10.1002/joc.1730http://dx.doi.org/10.1002/joc.1730http://dx.doi.org/10.1029/2005JD006290http://dx.doi.org/10.1029/2005JD006290http://dx.doi.org/10.1029/2005JD006290http://dx.doi.org/10.1029/2005JD006290http://dx.doi.org/10.1126/science.277.5324.364http://dx.doi.org/10.1126/science.277.5324.364http://dx.doi.org/10.1126/science.277.5324.364http://dx.doi.org/10.1126/science.277.5324.364http://dx.doi.org/10.1126/science.289.5487.2068http://dx.doi.org/10.1126/science.289.5487.2068http://dx.doi.org/10.1126/science.289.5487.2068http://dx.doi.org/10.3103/S1068373913040043http://dx.doi.org/10.3103/S1068373913040043http://dx.doi.org/10.3103/S1068373913040043http://dx.doi.org/10.3103/S1068373913040043http://dx.doi.org/10.3103/S1068373912020057http://dx.doi.org/10.3103/S1068373912020057http://dx.doi.org/10.3103/S1068373912020057http://dx.doi.org/10.3103/S1068373911060069http://dx.doi.org/10.3103/S1068373911060069http://dx.doi.org/10.3103/S1068373911060069http://dx.doi.org/10.3103/S1068373909060107http://dx.doi.org/10.3103/S1068373909060107http://dx.doi.org/10.3103/S1068373909060107http://dx.doi.org/10.3103/S1068373907010116http://dx.doi.org/10.3103/S1068373907010116http://dx.doi.org/10.3103/S1068373907010116http://dx.doi.org/10.3103/S1068373907010116http://dx.doi.org/10.3354/cr019193http://dx.doi.org/10.3354/cr019193http://dx.doi.org/10.3354/cr019193http://dx.doi.org/10.3354/cr019193

Nakamigawa, A., Ouprasitwong, N., Solofa, D., Tahani, L., Thuy, D.T., Tibig, L.,Trewin, B., Vediapan, K., Zhai, P., 2005. Change in mean temperature as apredictor of extreme temperature change in the Asia-Pacific region. Int. J.Climatol. 25, 1301–1330, http://dx.doi.org/10.1002/joc.1194.

Groisman, P.Y., Knight, R.W., Easterling, D.R., Karl, T.R., Hegerl, G.C., Razuvaev, V.N.,2005. Trends in intense precipitation in the climate record. J. Clim. 18, 1326–-1350, http://dx.doi.org/10.1175/JCLI3339.1.

Groisman, P.Y., Karl, T.R., Easterling, D.R., Knight, R.W., Jamason, P.F., Hennessy, K.J.,Suppiah, R., Page, C.M., Wibig, J., Fortuniak, K., Razuvaev, V.N., Douglas, A.,Førland, E., Zhai, P.-M., 1999. Changes in the probability of heavy precipitation:important indicators of climatic change. Clim. Change 43, 243–283, http://dx.doi.org/10.1023/A:1005432803188.

Haylock, M.R., Peterson, T.C., Alves, L.M., Ambrizzi, T., Anunciação, Y.M.T., Baez, J.,Barros, V.R., Berlato, M.A., Bidegain, M., Coronel, G., Corradi, V., Garcia, V.J.,Grimm, A.M., Karoly, D., Marengo, J.A., Marino, M.B., Moncunill, D.F., Nechet, D.,Quintana, J., Rebello, E., Rusticucci, M., Santos, J.L., Trebejo, I., Vincentu, L.A.,2006. Trends in total and extreme South American rainfall 1960-2000 and linkswith sea surface temperature. J. Clim. 19, 1490–1512, http://dx.doi.org/10.1175/JCLI3695.1.

IPCC, 2013. Summary for Policymakers. In: Stocker, T.F., Qin, D., Plattner, G.-K.,Tignor, M., Allen, S.K., Boschung, J., Nauels, A., Xia, Y., Bex, V., Midgley, P.M.(Eds.), Climate Change 2013: The Physical Science Basis. Contribution ofWorking Group I to the Fifth Assessment Report of the IntergovernmentalPanel on Climate Change. Cambridge University Press, Cambridge, UnitedKingdom; New York, New York, USA.

IPCC, 2012. Summary for Policymakers. In: Field, C.B., Barros, V., Stocker, T.F.,Qin, D., Dokken, D.J., Ebi, K.L., Mastrandrea, M.D., Mach, K.J., Plattner, G.-K.,Allen, S.K., Tignor, M., Midgley, P.M. (Eds.), Managing the Risks of ExtremeEvents and Disasters to Advance Climate Change Adaptation. CambridgeUniversity Press, Cambridge, United Kingdom; New York, USA (A Special Reportof Working Groups I and II of the Intergovernmental Panel on Climate Change).

IPCC, 2007. Summary for policymakers of climate change 2007: the physical sciencebasis. Contribution of Working Group I to the Fourth Assessment Report of theIntergovernmental Panel on Climate Change. Cambridge University Press,New York, USA.

Karl, T.R., Nicholls, N., Ghazi, A., 1999. CLIVAR/GCOS/WMOworkshop on indices andindicators for climate extremes: workshop summary. Clim. Change 42 (3–7), 10http://dx.doi.org/1023/A:1005491526870.

Katz, R.W., Brown, B.G., 1992. Extreme events in a changing climate-variability ismore important than averages. Clim. Change 21, 289–302, http://dx.doi.org/10.1007/BF00139728.

Kendall, M.G., 1975. Rank Correlation Methods. Charles Griffin, London, UK.Kharin, V., Zwiers, F.W., Zhang, X., Hegerl, G.C., 2007. Changes in temperature and

precipitation extremes in the IPCC ensemble of global coupled model simula-tions. J. Clim. 20, 1419–1444, http://dx.doi.org/10.1175/JCLI4066.1.

Klein Tank, A.M.G., Peterson, T.C., Quadir, D.A., Dorji, S., Zou, X., Tang, H., Santhosh, K.,Joshi, U.R., Jaswal, A.K., Kolli, R.K., Sikder, A.B., Deshpande, N.R., Revadekar, J.V.,Yeleuova, K., Vandasheva, S., Faleyeva, M., Gomboluudev, P., Budhathoki, K.P.,Hussain, A., Afzaal, M., Chandrapala, L., Anvar, H., Amanmurad, D., Asanova, V.S.,Jones, P.D., New, M.G., Spektorman, T., 2006. Changes in daily temperature andprecipitation extremes in central and south Asia. J. Geophys. Res. 111, D16105,http://dx.doi.org/10.1029/2005JD006316.

Klein Tank, A.M.G., Kon̈nen, G.P., 2003. Trends in indices of daily temperature andprecipitation extremes in Europe, 1946–99. J. Clim. 16, 3665–3680, http://dx.doi.org/10.1175/1520-0442(2003)016o3665:tiiodt42.0.co;2.

Kunkel, K.E., Roger, A.P., Stanley, A., 1999. Temporal fluctuations in weather andclimate extremes that cause economic and human health impacts: a review.Bull. Am. Meteorol. Soc. 80, 1077–1098, http://dx.doi.org/10.1175/1520-0477(1999)080o1077:TFIWAC42.0.CO;2.

Mann, H.B., 1945. Non-parametric tests against trend. Econometrica 13, 245–259.Manton, M.J., Della-Marta, P.M., Haylock, M.R., Hennessy, K.J., Nicholls, N.,

Chambers, L.E., Collins, D.A., Daw, G., Finet, A., Gunawan, D., Inape, K., Isobe,H., Kestin, T.S., Lefale, P., Leyu, C.H., Lwin, T., Maitrepierre, L., Ouprasitwong, N.,Page, C.M., Pahalad, J., Plummer, N., Salinger, M.J., Suppiah, R., Tran, V.L., Trewin,B., Tibig, I., Yee, D., 2001. Trends in extreme daily rainfall and temperature inSoutheast Asia and the South Pacific: 1961–1998. Int. J. Climatol. 21, 269–284,http://dx.doi.org/10.1002/joc.610.

Moberg, A., Jones, P.D., Lister, D., Walther, A., Alexander, L.V., Brunet, M., Chen, D.,Della-Marta, P.M., Jacobeit, J., Luterbacher, J., Yiou, P., Klein Tank, A.M.G.,Almarza, C., Auer, I., Barriendos, M., Bergstrom̈, H., Boḧm, R., Butler, J., Caesar,J., Drebs, A., Founda, D., Gerstengarbe, F.W., Giusi, M., Jónsson, T., Maugeri, M.,Österle, H., Pandzic, K., Petrakis, M., Srnec, L., Tolasz, R., Tuomenvirta, H.,Werner, P.C., Wanner, H., Xoplaki, E., 2006. Indices for daily temperature andprecipitation extremes in Europe analysed for the period 1901-2000.J. Geophys. Res. 111, D22106, http://dx.doi.org/10.1029/2006JD007103.

New, M., Hewitson, B., Stephenson, D.A., Tsiga, A., Kruger, A., Manhique, A., Gomez,B., Coelho, C.A.S., Masisi, D.N., Kululanga, E., Mbambalala, E., Adesina, F., Saleh,H., Kanyanga, J., Adosi, J., Bulane, L., Fortunata, L., Mdoka, M.L., Lajoie, R., 2006.Evidence of trends in daily climate extremes over Southern and West Africa.J. Geophys. Res. 111, D14102, http://dx.doi.org/10.1029/2005JD006289.

Nicholls, N., 2001. The insignificance of significance testing. Bull. Am. Meteorol. Soc.82, 981–986, http://dx.doi.org/10.1175/1520-0477(2001)082o0981:CAATIO42.3.CO;2.

Orlowsky, B., Seneviratne, S.I., 2012. Global changes in extremes events: regionaland seasonal dimension. Clim. Change 110, 669–696, http://dx.doi.org/10.1007/s10584-011-0122-9.

Page, C.M., Nicholls, N., Plummer, N., Trewin, B.C., Manton, M.J., Alexander, L.,Chambers, L.E., Choi, Y., Collins, D.A., Gosai, A., Della-Marta, P., Haylock, M.R.,Inape, K., Laurent, V., Maitrepierre, L., Makmur, E.E.P., Nakamigawa, H.,Ouprasitwong, N., McGree, S., Pahalad, J., Salinger, M.J., Tibig, L., Tran, T.D.,Vediapan, K., Zhai, P., 2004. Data rescue in the South-east Asia and South Pacificregion: challenges and opportunities. Bull. Am. Meteorol. Soc. 85, 1483–1489,http://dx.doi.org/10.1175/BAMS-85-10-1483.

Peterson, T.C., Taylor, M.A., Demeritte, R., Duncombe, D.L., Burton, S., Thompson, F.,Porter, A., Mercedes, M., Villegas, E., Fils, R.S., Klein-Tank, A.M.G., Martis, A.,Warner, R., Joyette, A., Mills, W., Alexander, L., Gleason, B., 2002. Recentchanges in climate extremes in the Caribbean region. J. Geophys. Res. 107, 4601,http://dx.doi.org/10.1029/2002JD002251.

Peterson, T.C., Folland, C., Gruza, G., Hogg, W., Mokssit, A., Plummer, N., 2001.Report on the activities of the Working Group on Climate Change Detection andRelated Rapporteurs 1998–2001. World Meteorological Organisation Rep.,Geneva, Switzerland p. 1071 (WCDMP-47, WMO-TD).

Peterson, T.C., Easterling, D.R., Karl, T.R., Groisman, P., Nicholls, N., Plummer, N.,Torok, S., Auer, I., Boḧm, R., Gullett, D., Vincent, L., Heino, R., Tuomenvirta, H.,Mestre, O., Szentimrey, T., Salinger, J., Førland, E.J., Hanssen-Bauer, I., Alex-andersson, H., Jones, P., Parker, D., 1998. Homogeneity adjustments of in situatmospheric climate data: a review. Int. J. Climatol. 18, 1493–1517, http://dx.doi.org/10.1002/(SICI)1097-0088(19981115)18:13o1493::AID-JOC32943.0.CO;2-T.

Russo, S., Sterl, A., 2011. Global changes in indices describing moderate temperatureextremes from the daily output of a climate model. J. Geophys. Res. 116,D03104, http://dx.doi.org/10.1029/2010JD014727.

Sen, P.K., 1968. Estimates of regression coefficient based on Kendall's tau. J. Am.Stat. Assoc. 63, 1379–1389, http://dx.doi.org/10.2307/2285891.

Sillmann, J., Kharin, V.V., Zhang, X., Zwiers, F.W., Bronaugh, D., 2013a. Climateextreme indices in the CMIP5 multi-model ensemble. Part 1: model evaluationin the present climate. J. Geophys. Res. 118, 1716–1733, http://dx.doi.org/10.1002/jgrd.50203.

Sillmann, J., Kharin, V.V., Zhang, X., Zwiers, F.W., Bronaugh, D., 2013b. Climateextreme indices in the CMIP5 multi-model ensemble. Part 2: future climateprojections. J. Geophys. Res. 118, 2473–2493, http://dx.doi.org/10.1002/jgrd.50188.

Skansi, M., Brunet, M., Sigró, J., Aguilar, E., Arevalo Groening, J.A., Bentancur, O.J.,Castellón Geier, Y.R., Correa Amaya, R.L., Jácome, H., Malheiros Ramos, A., OriaRojas, C., Pasten, A.M., Sallons Mitro, S., Villaroel Jiménez, C., Martínez, R.,Alexander, L.V., Jones, P.D., 2013. Warming and wetting signals emerging fromanalysis of changes in climate extreme indices over South America. GlobalPlanet. Change 100, 295–307, http://dx.doi.org/10.1016/j.gloplacha.2012.11.004.

Tebaldi, C., Hayhoe, K., Arblaster, J.M., Meehl, G.A., 2006. Going to the extremes: anintercomparison of model simulated historical and future changes in extremeevents. Clim. Change 79, 185–211, http://dx.doi.org/10.1007/s10584-006-9051-4.

Wang, X.L., Feng, Y., 2010a. RHtestsV3 User Manual. Climate Research Division.Science and Technology Branch, Environment Canada p. 26.

Wang, X.L., Chen, H., Wu, Y., Feng, Y., Pu, Q., 2010b. New techniques for detectionand adjustment of shifts in daily precipitation data series. J. Appl. Meteorol.Climatol. 49, 2416–2436, http://dx.doi.org/10.1175/2010JAMC2376.1.

Wang, X.L., 2008a. Accounting for autocorrelation in detecting mean-shifts inclimate data series using the penalized maximal t or F test. J. Appl. Meteorol.Climatol. 47, 2423–2444, http://dx.doi.org/10.1175/2008JAMC1741.1.

Wang, X.L., 2008b. Penalized maximal F-test for detecting undocumented mean-shifts without trend-change. J. Atmos. Oceanic Techol. 25, 368–384, http://dx.doi.org/10.1175/2007/JTECHA982.1.

World Bank, 2006. Drought: Management and Mitigation Assessment for CentralAsia and the Caucasus. World Bank – Regional and Country Profiles andStrategies, Washington D.C., USA.

Zhang, X., Alexander, L.V., Hegerl, G.C., Klein-Tank, A., Peterson, T.C., Trewin, B.,Zwiers, F.W., 2011. Indices for monitoring changes in extremes based on dailytemperature and precipitation data. Wiley Interdiscip. Rev.: Clim. Change 2,851–870, http://dx.doi.org/10.1002/wcc.147.

Zhang, X., Hegerl, G., Zwiers, F., Kenyon, J., 2005a. Avoiding inhomogeneity inpercentile-based indices of temperature extremes. J. Clim. 18, 1641–1651, http://dx.doi.org/10.1175/JCLI3366.1.

Zhang, X., Aguilar, E., Sensoy, S., Melkonyan, H., Tagiyeva, U., Ahmed, N., Kutaladze,N., Rahimzadeh, F., Taghipour, A., Hantosh, T.H., Albert, P., Semawi, M., 2005b.Trends in Middle East climate extreme indices from 1950 to 2003. J. Geophys.Res. 110, D22104, http://dx.doi.org/10.1029/2005JD006181.

Zhang, X., Yang, F., 2004. RClimDex (1.0) User Guide. Climate Research BranchEnvironment Canada, Downsview (Ontario, Canada) p. 22.

Zhang, X., Vincent, L.A., Hogg, W.D., Niitsoo, A., 2000. Temperature and precipita-tion trends in Canada during the 20th century. Atmos. Ocean 38, 395–429,http://dx.doi.org/10.1080/07055900.2000.9649654.

I. Keggenhoff et al. / Weather and Climate Extremes 4 (2014) 75–85 85

http://dx.doi.org/10.1002/joc.1194http://dx.doi.org/10.1002/joc.1194http://dx.doi.org/10.1002/joc.1194http://dx.doi.org/10.1175/JCLI3339.1http://dx.doi.org/10.1175/JCLI3339.1http://dx.doi.org/10.1175/JCLI3339.1http://dx.doi.org/10.1023/A:1005432803188http://dx.doi.org/10.1023/A:1005432803188http://dx.doi.org/10.1023/A:1005432803188http://dx.doi.org/10.1023/A:1005432803188http://dx.doi.org/10.1175/JCLI3695.1http://dx.doi.org/10.1175/JCLI3695.1http://dx.doi.org/10.1175/JCLI3695.1http://dx.doi.org/10.1175/JCLI3695.1http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref18http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref18http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref18http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref18http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref18http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref18http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref19http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref19http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref19http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref19http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref19http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref19http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref20http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref20http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref20http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref20http://dx.doi.org/1023/A:1005491526870http://dx.doi.org/10.1007/BF00139728http://dx.doi.org/10.1007/BF00139728http://dx.doi.org/10.1007/BF00139728http://dx.doi.org/10.1007/BF00139728http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref23http://dx.doi.org/10.1175/JCLI4066.1http://dx.doi.org/10.1175/JCLI4066.1http://dx.doi.org/10.1175/JCLI4066.1http://dx.doi.org/10.1029/2005JD006316http://dx.doi.org/10.1029/2005JD006316http://dx.doi.org/10.1029/2005JD006316http://dx.doi.org/10.1175/1520-0442(2003)016<3665:tiiodt>2.0.co;2http://dx.doi.org/10.1175/1520-0442(2003)016<3665:tiiodt>2.0.co;2http://dx.doi.org/10.1175/1520-0442(2003)016<3665:tiiodt>2.0.co;2http://dx.doi.org/10.1175/1520-0442(2003)016<3665:tiiodt>2.0.co;2http://dx.doi.org/10.1175/1520-0442(2003)016<3665:tiiodt>2.0.co;2http://dx.doi.org/10.1175/1520-0442(2003)016<3665:tiiodt>2.0.co;2http://dx.doi.org/10.1175/1520-0442(2003)016<3665:tiiodt>2.0.co;2http://dx.doi.org/10.1175/1520-0442(2003)016<3665:tiiodt>2.0.co;2http://dx.doi.org/10.1175/1520-0477(1999)080<1077:TFIWAC>2.0.CO;2http://dx.doi.org/10.1175/1520-0477(1999)080<1077:TFIWAC>2.0.CO;2http://dx.doi.org/10.1175/1520-0477(1999)080<1077:TFIWAC>2.0.CO;2http://dx.doi.org/10.1175/1520-0477(1999)080<1077:TFIWAC>2.0.CO;2http://dx.doi.org/10.1175/1520-0477(1999)080<1077:TFIWAC>2.0.CO;2http://dx.doi.org/10.1175/1520-0477(1999)080<1077:TFIWAC>2.0.CO;2http://dx.doi.org/10.1175/1520-0477(1999)080<1077:TFIWAC>2.0.CO;2http://dx.doi.org/10.1175/1520-0477(1999)080<1077:TFIWAC>2.0.CO;2http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref28http://dx.doi.org/10.1002/joc.610http://dx.doi.org/10.1002/joc.610http://dx.doi.org/10.1002/joc.610http://dx.doi.org/10.1029/2006JD007103http://dx.doi.org/10.1029/2006JD007103http://dx.doi.org/10.1029/2006JD007103http://dx.doi.org/10.1029/2005JD006289http://dx.doi.org/10.1029/2005JD006289http://dx.doi.org/10.1029/2005JD006289http://dx.doi.org/10.1175/1520-0477(2001)082<0981:CAATIO>2.3.CO;2http://dx.doi.org/10.1175/1520-0477(2001)082<0981:CAATIO>2.3.CO;2http://dx.doi.org/10.1175/1520-0477(2001)082<0981:CAATIO>2.3.CO;2http://dx.doi.org/10.1175/1520-0477(2001)082<0981:CAATIO>2.3.CO;2http://dx.doi.org/10.1175/1520-0477(2001)082<0981:CAATIO>2.3.CO;2http://dx.doi.org/10.1175/1520-0477(2001)082<0981:CAATIO>2.3.CO;2http://dx.doi.org/10.1175/1520-0477(2001)082<0981:CAATIO>2.3.CO;2http://dx.doi.org/10.1175/1520-0477(2001)082<0981:CAATIO>2.3.CO;2http://dx.doi.org/10.1007/s10584-011-0122-9http://dx.doi.org/10.1007/s10584-011-0122-9http://dx.doi.org/10.1007/s10584-011-0122-9http://dx.doi.org/10.1007/s10584-011-0122-9http://dx.doi.org/10.1175/BAMS-85-10-1483http://dx.doi.org/10.1175/BAMS-85-10-1483http://dx.doi.org/10.1175/BAMS-85-10-1483http://dx.doi.org/10.1029/2002JD002251http://dx.doi.org/10.1029/2002JD002251http://dx.doi.org/10.1029/2002JD002251http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref36http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref36http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref36http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref36http://dx.doi.org/10.1002/(SICI)1097-0088(19981115)18:13<1493::AID-JOC329>3.0.CO;2-Thttp://dx.doi.org/10.1002/(SICI)1097-0088(19981115)18:13<1493::AID-JOC329>3.0.CO;2-Thttp://dx.doi.org/10.1002/(SICI)1097-0088(19981115)18:13<1493::AID-JOC329>3.0.CO;2-Thttp://dx.doi.org/10.1002/(SICI)1097-0088(19981115)18:13<1493::AID-JOC329>3.0.CO;2-Thttp://dx.doi.org/10.1002/(SICI)1097-0088(19981115)18:13<1493::AID-JOC329>3.0.CO;2-Thttp://dx.doi.org/10.1002/(SICI)1097-0088(19981115)18:13<1493::AID-JOC329>3.0.CO;2-Thttp://dx.doi.org/10.1002/(SICI)1097-0088(19981115)18:13<1493::AID-JOC329>3.0.CO;2-Thttp://dx.doi.org/10.1002/(SICI)1097-0088(19981115)18:13<1493::AID-JOC329>3.0.CO;2-Thttp://dx.doi.org/10.1002/(SICI)1097-0088(19981115)18:13<1493::AID-JOC329>3.0.CO;2-Thttp://dx.doi.org/10.1029/2010JD014727http://dx.doi.org/10.1029/2010JD014727http://dx.doi.org/10.1029/2010JD014727http://dx.doi.org/10.2307/2285891http://dx.doi.org/10.2307/2285891http://dx.doi.org/10.2307/2285891http://dx.doi.org/10.1002/jgrd.50203http://dx.doi.org/10.1002/jgrd.50203http://dx.doi.org/10.1002/jgrd.50203http://dx.doi.org/10.1002/jgrd.50203http://dx.doi.org/10.1002/jgrd.50188http://dx.doi.org/10.1002/jgrd.50188http://dx.doi.org/10.1002/jgrd.50188http://dx.doi.org/10.1002/jgrd.50188http://dx.doi.org/10.1016/j.gloplacha.2012.11.004http://dx.doi.org/10.1016/j.gloplacha.2012.11.004http://dx.doi.org/10.1016/j.gloplacha.2012.11.004http://dx.doi.org/10.1007/s10584-006-9051-4http://dx.doi.org/10.1007/s10584-006-9051-4http://dx.doi.org/10.1007/s10584-006-9051-4http://dx.doi.org/10.1007/s10584-006-9051-4http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref44http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref44http://dx.doi.org/10.1175/2010JAMC2376.1http://dx.doi.org/10.1175/2010JAMC2376.1http://dx.doi.org/10.1175/2010JAMC2376.1http://dx.doi.org/10.1175/2008JAMC1741.1http://dx.doi.org/10.1175/2008JAMC1741.1http://dx.doi.org/10.1175/2008JAMC1741.1http://dx.doi.org/10.1175/2007/JTECHA982.1http://dx.doi.org/10.1175/2007/JTECHA982.1http://dx.doi.org/10.1175/2007/JTECHA982.1http://dx.doi.org/10.1175/2007/JTECHA982.1http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref49http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref49http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref49http://dx.doi.org/10.1002/wcc.147http://dx.doi.org/10.1002/wcc.147http://dx.doi.org/10.1002/wcc.147http://dx.doi.org/10.1175/JCLI3366.1http://dx.doi.org/10.1175/JCLI3366.1http://dx.doi.org/10.1175/JCLI3366.1http://dx.doi.org/10.1175/JCLI3366.1http://dx.doi.org/10.1029/2005JD006181http://dx.doi.org/10.1029/2005JD006181http://dx.doi.org/10.1029/2005JD006181http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref53http://refhub.elsevier.com/S2212-0947(14)00033-4/sbref53http://dx.doi.org/10.1080/07055900.2000.9649654http://dx.doi.org/10.1080/07055900.2000.9649654http://dx.doi.org/10.1080/07055900.2000.9649654

Trends in daily temperature and precipitation extremes over Georgia, 1971–2010IntroductionData and methodsStudy areaData, quality control and homogeneity testingClimate extreme indices and analytical methods

Results and discussionTrends in temperature extremesTrends in precipitation extremes

ConclusionsAcknowledgmentsAppendixReferences